Nominal shocks and relative price variability: an empirical study for the Peruvian economy

Abstract

Inflation has been a recurrent and critical problem in many Latin American countries. Inflation is often combined with, among other problems, serious distortions in the structure of relative prices thus reducing efficiency of the pricing mechanism in allocating resources.

The purpose of this study is to examine the effects of inflation and other variables on relative price variability in Peru using two different types of models. After preliminary evaluation of the stationarity properties of the series, a relative price variability measure is constructed using monthly data on 32 components of the CPI over the period 1979:12-1988:07. For the first models, series of expected and unexpected inflation in Peru, real exchange rate movements, and U.S. relative price variability are constructed and the effects of these variables on observed relative price variability are determined. The results indicate that increasing levels and unpredictability of inflation cause increased dispersion of relative prices.

A distinction between expected and unexpected relative price changes is made in the second model. This distinction is relevant because not all price movements are viewed by agents as surprises that confuse price signals. To account for this distinction, a measure of conditional relative price variability is estimated using Engle's (1982) autoregressive conditional heteroskedasticity approach. Similarly, the conditional variance of domestic inflation is estimated and used as a measure of price uncertainty.

Effects of the time-dependent conditional variances of domestic inflation and real exchange rate on a weighted average of the relative price shocks normalized by their conditional variances are evaluated and found statistically significant. Finally, both methods of testing empirically the hypothesis that inflation uncertainty increases relative price variability provided consistent results.